How do I test whether a habit is actually working for me?
Measure yourself for a week or two before you start the habit, then start it and keep measuring the same thing the same way. The comparison that tells you anything is your own baseline against your own intervention period — not how you feel now against how you remember feeling, which is unreliable because your memory of the earlier period is reconstructed after you already believe the habit works. Stopping the habit for a week and watching whether the effect reverses is the single strongest check available to one person, because almost nothing else produces that pattern by chance.
The four steps
- Pick one measure and write down how you will score it. One number, once a day, same time. "Energy at 3pm, 1 to 10" beats "how did the day go" because you can repeat it identically for six weeks.
- Take a baseline for one to two weeks before changing anything. This is the step people skip, and skipping it makes everything after it uninterpretable.
- Start the habit and keep the measure identical. Change nothing else you can control in the same window.
- Stop for a week, then start again. If the measure follows the habit both times, you have something. If it does not move, you have saved yourself months.
Why the baseline is the whole thing
Without a baseline you are comparing today against a memory, and the memory was formed after you decided the habit was a good idea. You will find the effect you expected. That is not a character flaw or a failure of effort — it is what reconstruction does, and it happens to researchers too, which is why the design exists.
A written baseline is a record made by someone who did not yet know the answer. That is the only version of you worth arguing with.
What makes a result meaningless
- Changing three things at once. If you started running, cut caffeine, and moved your bedtime in the same week, no outcome tells you which one did it.
- Changing the measure partway through. A scale you redefine is a scale you cannot compare across time.
- Testing something whose effect is slower than your test. Two weeks will not settle a question about strength training or antidepressant response.
- Stopping the moment it looks good. The early days of any change include the effect of having decided to change, which fades.
How far this goes
This is a real method, not a self-help exercise. In medicine it is called an n-of-1 trial, and the US Agency for Healthcare Research and Quality publishes a full design guide for running them on single patients. What you are doing at home is a looser version: no blinding, no randomised order, one measure. It is weaker than the clinical version and much stronger than deciding by memory.
Where Emerge fits
Emerge has this built in as Your Lab: you pick one thing to change and one thing to measure, it holds you to a baseline period before the change starts, and at the end it gives you the verdict the data supports — including 'this did nothing for you', which is the result the method exists to be able to return.
Emerge on the App Store →Related questions
How long should I test a habit before deciding?
Longer than feels necessary. In Lally et al. (2010), the median time for a daily behaviour to become automatic was 66 days, with a range from 18 to 254 across participants. A two-week test can detect an immediate effect on mood or energy; it cannot tell you whether a habit will stick.
Do I need an app to run a self-experiment?
No. A note on your phone with a date and one number a day is a complete instrument. An app helps with two things only: reminding you at the same time each day, and stopping you from reading the result before the period ends.
Sources
Last updated 2026-09.